I've decided to launch my own consulting firm called Sanger Consulting. Read my blog post to learn more.

Do you run a news startup that features original reporting—not just news aggregated from other sources? Do you want to make it a standout, the sort of service people will point to as “the best solution to all the problems with fake news”? Then I want to give you some free advice. And look in the comments to see if anyone agrees that this is a good idea. Maybe not. We will see.

No sensible reader accepts the claims of journalists without question anymore, if ever they did. Considering their track record, especially in recent years, this is only appropriate. Journalists might not like this state of affairs, but it is not likely to go away anytime soon. The reason we are more aware of journalistic malpractice today than ever before is that sources are often as close as the reporters who report on them. Besides, reporters wear their bias on their sleeve. It follows that responsible news consumption must be critical.

It follows that we need new and better tools to decide whether various claims made are actually accurate. And that is really the root of the insight I want to give you:

Stop thinking of journalism as writing a story. Start thinking of it as sharing research notes.

What does this mean?

Well, what is the purpose of journalism in the first place? I propose to back to basics: let us set aside the ideological purposes of journalism and think about the more essential function or activity, both from the reporter’s and from the reader’s perspective. The reporter is trying to introduce the reader to some new facts as accurately and efficiently, preferably in a narrative that captures the reader’s interest, perhaps while advancing an editorial angle (not that I approve of the latter). Surely, as old school journalists think, that is enough.

The reader is indeed trying to apprise himself of new facts. But he has to negotiate at least three issues that are at the forefront of his mind:

  1. Importance: How important is this news, really? What might make it important, if it is? Has it been blown out of proportion?
  2. Credibility: Is the purported news credible? Is there adequate evidence, especially in terms of sourcing, to believe it? What is the confidence level of the news?
  3. Neutrality: Are there important facts being left out? Would other sources disagree?

I do not think reporters dwell much on these needs of the reader. For one thing, standard journalistic practice in 2020 is fixated on creating narrative, as if they were novelists or something. In so doing, a reporter might give enough background necessary to appreciate the importance of the story; it might say enough about sources to establish the credibility of the information; it might lay out the context sufficiently that the reader can appreciate that a complete, appropriate range of views are being presented in a way that invites the reader to make up his own mind.

But these days, most news stories do none of this. Stories are presented as if they were all uniformly important, for often undisclosed reasons. Articles tend to lead with sensational claims, without considering why sources (or journalists) might exaggerate. Sources are frequently anonymous, and even when a source is fully identified, key information—specifically why they are in a position to know—is hidden or obscured. The most relevant scientific data, video evidence, documents, etc., are only sometimes included. As to neutrality, that has mostly gone the way of the dodo; but it is still something readers desperately want. This is not a quirk. Responsible readers know they must have the whole story, told from all sides, if they are to come to a rational opinion about it. Otherwise they know they are probably being manipulated.

Imagine, instead of this sort of hit-or-miss news article, readers were presented a news research summary, which featured:

  • Bullet points summarizing the main facts, in order of importance (together with sources).
  • A collection of key quotations from source interviews and documents.
  • A list of sources (people), with their most specific relevant qualifications in the relevant subject matter fully stated, at the top of the article. If a source is anonymous, then as much information as possible (such as that a person is “a retired U.S. intelligence official”).
  • A list of any available supporting media, including images, videos, links to or copies of documents and polls, etc. A nice subsection would be a list of relevant tweets and other social media sources (assuming they are newsworthy).
  • A list of previous reporting consulted.
  • A background narrative or, perhaps better, answers to likely questions about matters necessary to understand in order to appreciate the news, its importance, relevant controversies, etc.
  • Recent commentary on the issues involved, scrupulously divided into competing sides as necessary.
  • Now, if in addition to this you wanted to have a traditional news narrative, that would be OK, as a “nice-to-have”; but I suspect that, once all these other resources were marshaled, the narrative would be regarded as a boring aside that is better skipped.

In short, what we really need is a resource that we can use to research and come to our own conclusions. Of course, a news research summary could be made just as biased as traditional journalism; but if done well, it would be exactly the sort of thing professional fact-checkers need to do their work. In any event, a startup that developed such resource pages should use trial and error to determine precisely the best format; it is not especially clear to me what the very best format would look like. Certain kinds of reports would have specific requirements; for example, a report about a poll should include a section, with both summary and details, about polling methodology and about the pollster.

If done well, I and I think many people would be willing to put down good money for such reporting.

Theoretically, this could be crowdsourced, e.g., on a wiki, but offhand I am inclined to think a crowdsourced version would not work out very well for the simple reason that people do not generally volunteer to do difficult gruntwork, and compiling this sort of resource would be quite difficult work indeed. But maybe; I could be wrong.